22 Jan 2010

Prefixes in Minoan

I've read many people writing of 'prefixes' in Minoan but I frown on these notions. Yes, I know, I'm a party-pooper but I don't like things that make no sense. Knowing nothing else about this undeciphered language (and believe me, no one really knows much on this language unfortunately), it should be understood that 90% of world languages are either SOV or SVO[1]. Of the two, SOV is the most common word order overall and with this word order comes the strong tendency towards suffixing with very few prefixes.[2] For example, in Proto-Indo-European the only prefix I can think of offhand that is securely reconstructed is privative *n̥- (nb. the past augment *h₁e- was only dialectal variation). The fact that prefixing languages are less common than suffixing ones should alone be enough to dissuade people into looking for genuine prefixes in this language until everything else is better understood.

I'm happy to note that Andras Zeke finally resurfaced with his latest post, More on the libation formula - the statuette of Poros Herakliou. While it's another interesting post full of fresh points of view, I have to comment on the idea that, concerning U-QE-TI, "the 'U'-prosthetics might have been verbal formatives in the Minoan language". John Younger too mentions prefixes, citing Yves Duhoux[3].

Given the statistics of world language patterns and if we know next to nothing about Minoan, leaping to the assumption that we should find prefixes in this language is an attempt to grasp at straws. We need to focus on the more likely, not the least likely! I can easily explain away all of these alleged 'prefixes' that I've so far seen identified by others. Prefixes are simply unnecessary and only obstruct decipherment:
  • I-/YA-
    The YA-/A- alternation in Minoan is merely a reflection of the phonotactically motivated avoidance of word-initial semivowel /j/. The same constraint was coincidentally in effect in Etruscan. The differences seen between YA-SA-SA-RA-ME (IO Za 6) and A-SA-SA-RA-ME (PK Za 11) then don't involve morphology but rather spelling preferences and there is then no need to define a function for this 'prefix' that's in effect imaginary. The related I-prefix that is supposed is effectively only demonstrated by a single pair, I-DA-MA-TE (AR Zf 1, Zf 2) versus DA-MA-TE (KY Za 2), but isn't sufficient evidence since a single spelling aberration can't rule out other possibilities like syncope or scribal error.
  • U-
    No such suffix. It must be first established by identifying pairs with and without the morpheme. The argument that "words beginning with U- are relatively rare" isn't enough in establishing such a prefix.
  • TAN-
    This is surely not a prefix. It's a preposed distal demonstrative *tan, declined in the accusative case and identical to that found in Etruscan. Thus TA-NA-SU-TE-KE (PR Za 1) is two words, *tan suteke, the direct object of the phrase in question.
So I thus far deny all prefixes in Minoan.


NOTES
[1] Blake, Case (2001), p.99 (see link).
[2] Fuss, The Rise of Agreement: A Formal Approach to the Syntax and Grammaticalization of Verbal Inflection (2005), p.153 (see link).
[3] Linear A Texts: Homepage - 13c. Prefixes.

5 comments:

  1. What about the "s-mobile" in PIE? Why aren't you counting that as a prefix?
    Also, about phonotactically avoided /j/ in Etruscan: do you mean like in "Larθal" vs. "Larθial", or "Uni" vs. Roman "Juno"?

    ReplyDelete
  2. Seadog Driftwood: "What about the "s-mobile" in PIE? Why aren't you counting that as a prefix?"

    The *s-mobile has no proven function in PIE and therefore can't be called a genuine grammatical prefix.

    I wager it's merely a petrifact from Neolithic loans derived from Proto-Semitic where *s- served as a causative marker. Proto-Semitic was coincidently VSO, not SOV.

    And to add, the reason why the privative *n̥- was prefixed was probably because the negative adverb from which it derives, *ne, was likewise preposed to the verb, the reason for which I've already explained before in Negational particles, negational verbs and negational adverbs.

    Hence, PIE, like the majority of SOV languages around the world, have few to no prefixes.


    "Also, about phonotactically avoided /j/ in Etruscan: do you mean like in 'Larθal' vs. 'Larθial', or 'Uni' vs. Roman 'Juno'?"

    You overlooked what I already clearly wrote: "avoidance of word-initial semivowel /j/." And, Iūnō is the *Latin* name; Uni is the Etruscan which, lo and behold, avoids the word-initial semivowel /j/. Case in point.

    ReplyDelete
  3. >"Also, about phonotactically avoided /j/ in Etruscan: do you mean like in 'Larθal' vs. 'Larθial', or 'Uni' vs. Roman 'Juno'?"

    "You overlooked what I already clearly wrote: "avoidance of word-initial semivowel /j/." And, Iūnō is the *Latin* name; Uni is the Etruscan which, lo and behold, avoids the word-initial semivowel /j/. Case in point."

    >.< Argh twice over on my part! But actually, the /j/ onset of iūnō was what I was trying to point out.

    As for the paucity of prefixes, I certainly agree that were are indeed few in PIE. I'm not sure if reduplication counts, though.

    On a semi-related topic, the most recent post on the Minoan Language Blog, the identification of TA as *tha in the instances described, especially in "HT25: A-RI-NI-TA (*arn-intha?)"- what do you make of that? And if it's plausibly (with many an "if", obviously!) accurate, what might it tell about the pronunciation of /t/ in Minoan, at least in the circumstances described in the post?

    ReplyDelete
  4. Seadog Driftwood: "I'm not sure if reduplication counts, though."

    Nope. Prefixation requires a definable prefix, obviously, but reduplication is wholly dependent on the shape of the root from which it's derived. Every PIE verb has varying reduplicated elements (eg. *pipeh₃-, *(s)testoh₂-, etc.). I'd define this phenomenon not so much as "prefixation" as "internal restructuring".

    "On a semi-related topic, the most recent post on the Minoan Language Blog, the identification of TA as *tha in the instances described, especially in "HT25: A-RI-NI-TA (*arn-intha?)"- what do you make of that?"

    It's certainly possible that Linear A didn't represent aspiration contrasts that were in the language. If Minoan is related to Etruscan, we'd either expect Proto-Aegean *t and *tʰ to fall together as /t/ or at the very least be represented both with the same t-series in writing. This reasoning goes as well for the other plosives, *k/*kʰ and *p/*pʰ.

    ReplyDelete
  5. Hi, Glen! I've been following your comments on Andras' blog with interest, and finally saw a link to your blog today. Mine is http://minoan.deaditerranean.com

    RE: A-/JA-

    I agree that this is likely a spelling preference... I tend to think of it a bit like the Greek accent which can add a soft "h" breathing in front of an alpha. Especially the A since it has been shown to be both present and absent as a prefix to several sign groups.

    ReplyDelete